Meeting Abstract

P1.101  Tuesday, Jan. 4  A digital simulation of a carangiform fish: self-propulsion and autonomous navigation PHAM, Tong T*; DAHAL, Bidur K; LIN, Khine; LIEW, Chun W; ROOT, Robert G; LONG JR, John H; Lafayette College; Lafayette College; Lafayette College; Lafayette College; Lafayette College; Vassar College liew@cs.lafayette.edu

Physics-based simulations can be an effective tool to test hypotheses regarding the function and evolution of certain morphological features. In this study, we explore how the backbones of carangiform fish might have evolved. We constructed a two-dimensional model of the Tunicate Tadpole larvae. In our previous work on self-propelled swimmers, we used Lighthill’s slender body theory to model the thrust. Unfortunately, this theory is invalid for slow velocities found during maneuvers such as starting, breaking, and turning. Thus, we switched to use Childress’ model (1977), which has lower bounds of applicability with regards to velocity and hence suitable for our cases. Our genetic algorithm used a fitness function that favored a combination of velocity, time spent, distance from, and stability of orbit around the food source. With respect to navigation prowness, results showed a strongly negative correlation for tail length, and a weakly negative correlation for tail stiffness. This shows that a backbone conferred a degree of adaptive advantage in foraging by increasing maneuverability. This work was supported by NSF DBI-0442269.